The modern world is increasingly dependent upon artificial intelligence and machine learning. Applications range from decision-making in areas such as health and finance to autonomous vehicle control. However, current deep-learning machines and neural network algorithms have important limitations, namely ineffective learning rules, long training, and high power consumption (1). In the present project, we aim to address these limitations by using the human nervous system itself as a model, which can process external information in a power-efficient way. To accomplish this objective, we utilize cortical neuronal networks derived from human induced pluripotent stem cells, as well as primary cultures comprising neurons and glial cells. These cultures are cultivated on a high-density multielectrode array (hdMEA) chip, and high-resolution recordings of neuronal activity have been performed. Moreover, our setup integrates a modular microfluidics device with the hdMEA chip, which precisely controls the formation of a neuronal network. Our results show that with the inclusion of modular microfluidics devices, we obtain a more brain-like complex functional activity. Additionally, precise bidirectional microelectrode stimulation is applied to specific cells within the network. The results reveal that our set-up induces plasticity changes in the neuronal cultures. Analysis of parameters governing neuronal circuit responses to patterned stimuli paves the way toward trainable circuits with desired processing capabilities. The acquired data provides detailed information about the complex network dynamics of large populations of human neurons, facilitating advancements in neuroinformatics, cognitive neuroscience, and various related fields.­­­(1) E. Strubell, et al., https://arxiv.org/abs/1906.02243v1 (2019)